ENGINEERING AND PROCESSING
Effect of Pump Rate and Table Slope on Starch Recovery for
a 100-g Laboratory Wet-Milling Procedure
S. K. SINGH' and S. R. ECKHOFF 2
ABSTRACT
Cereal Chem. 73(1):51-53
A replicated full-factorial experiment with four starch table slopes
(0.0052, 0.0104, 0.0156, and 0.0208 cm/cm) and five pump rates (40,
45, 50, 55, and 60 ml/min) was used to determine the relationship
between the table parameters of slope and slurry pumping rate and
starch yield and the protein content of starch for a 2.44-m x 5.08-cm
aluminum channel used as a starch table in a 100-g laboratory wet-milling procedure. The tests showed that both starch yield and the protein
content in the starch decreased with increasing table slope and pumping
rate. The rate of starch yield loss with increasing table slope and pumping rate was approximately linear. However, protein content of starch
was relatively unaffected by table slope at slopes >0.0104 cm/cm. The
lowest table slope (0.0052) had significantly more protein in the starch
than did the other three table slopes tested. Selection of the appropriate
table slope and pumping rate for use with the 100-g laboratory wetmilling procedure was based upon starch yield and protein in starch
results from tabling the same starch slurry on an 8.3-cm x 6.1-im table as
performed in a 1-kg laboratory wet-milling procedure. A table slope of
0.0104 cm/cm and a pumping rate of 50 ml/min was the combination
that gave the starch yield (81.8 vs. 80.0%) and protein content in starch
(0.50 vs. 0.52%) closest to those of the 1-kg tabling procedure.
The density separation of starch and protein in the corn wetmilling process is currently accomplished using a series of continuous centrifuges and hydrocyclones (Watson 1984, Blanchard
1993). However, up until approximately 40 years ago, corn starch
was separated from protein in industrial processes by pumping
the starch slurry onto sloped wooden troughs called starch tables.
Starch would settle on the table, while the protein remained suspended in the water and would be carried off the end of the table.
The tables were typically 0.6-m wide and 40-m long with a slope
of =0.0052 m/mr (Kerr 1950). Factors affecting the performance of
the industrial starch tables included the table slope, specific gravity of the starch-protein slurry, temperature of the slurry, table
length, and pump volume per unit width of table (Berkhout 1976).
Three methods have been used in laboratory and pilot-size wet
milling to separate starch from protein: 1) batch centrifugation
(Hassanean and Abdel-Wahed 1986, Steinke and Johnson 1991);
2) hydrocyclone (Rubins 1990, Singh and Eckhoff 1995); and 3)
starch tabling (Watson et al 1951; Eckhoff et al 1993; Pelshenke
and Lindemann 1954; Anderson 1957, 1963; Weller 1987;
Wehling et al 1993). There has been no direct comparison
between the three procedures, although Singh and Eckhoff (1995)
compared the hydrocyclone with the starch table. They found that
at the conditions they operated the hydrocyclone, starch yield was
reduced and protein content in the starch was higher for starch
recovered by the hydrocyclone procedure. They felt that further
refinement of the procedure, including recycling of the protein
overflow, would yield acceptable results. Results from the references above and other literature (Eckhoff and Tso 1991, Steinke
et al 1991, Fox et al 1992, Wang and Johnson 1992, Fox and
Eckhoff 1993, Rausch et al 1993, Shandera et al 1995) indicate
that all three methods appear to be capable of giving starch yields
near industrial levels with low starch protein content and with low
standard deviation between replicates, although not all research
studies achieved the same level of accuracy and precision.
Each of the three methods appear to have different advantages.
The batch centrifugation and the hydrocyclone are faster than
tabling, but tabling has fewer subjective steps in its operation and
requires less expensive equipment. Centrifugation and use of a
hydrocyclone more closely emulate the current separation technology in the industry, take up less floor space, and are more
adjustable for working with hard-to-separate starch samples.
However, the hydrocyclone does not appear to be adaptable to
sample sizes much less than 1 kg.
Pelshenke and Lindemann (1954) used a short starch table to
separate starch and protein following a two-stage centrifugal
washing procedure for a 100-200 g sample procedure. They did
not report basic information on the slope of the table, the slurry
flow rate, or the specific gravity of the slurry, but they were able
to achieve good separation as indicated by a high starch yield,
low protein content in the starch, and a high degree of precision
in measuring starch yield.
In developing a laboratory wet-milling procedure that uses a
100-g sample size, Eckhoff et al (1996) used a 2.44-m x 5.08-cm
aluminum channel as a starch table. There is no scientific data
available on the relationship between the various parameters for
tabling such as slurry temperature, slurry specific gravity, table
length, table slope, and slurry pumping rate. In laboratory practice, it is possible to hold slurry temperature, slurry specific gravity, and table length constant. It is necessary to know the relationship between the starch quality and starch yield and the table
slope and slurry pumping rate. The objective of this research was
to determine the relationship between table slope, the pumping
rate onto the table, and the resulting starch yield and starch protein content for a 2.44-m x 5.08-cm aluminum starch table to be
used for starch-protein separation in a 100-g laboratory wet-milling
procedure.
Former graduate research assistant, University of Illinois. Current address:
Department of Food Science and Human Nutrition, Iowa State University, Ames,
IA 50011.
2
Professor, Department of Agricultural Engineering, University of Illinois,
Urbana, IL 61801. Corresponding author. E-mail: [email protected]
Publication no. C-1996-0109-02R.
O 1996 American Association of Cereal Chemists, Inc.
METHODS AND MATERIALS
A quantity of starch-protein slurry was obtained, all at one
time, from the feed to the primary centrifuge, in a wet-milling
facility. Commercial starch-protein slurry was utilized in the test
to minimize variability that would occur from repetitive
laboratory wet milling to produce starch-protein slurry. The slurry
was refrigerated to prevent deterioration of the starch until the
random duplicated tests were performed. When running a test
condition, the slurry was well mixed, a representative sample
withdrawn, warmed to room temperature, and adjusted to a 1.04
Vol. 73, No. 1,1996
51
specific gravity (=9.8% solids) by the addition of distilled water
before tabling. Three 100-ml samples of the slurry were used for
solid determination by the two-stage air-drying procedure (AACC
1983).
Samples (1,000 ml) of the prepared mill starch with a specific
gravity of 1.04 were tabled at each of 20 combinations of table
slope and pump rate in a duplicated full-factorial experimental
design comparing four table slopes (0.0052, 0.0104, 0.0156, and
0.0208 cm/cm) and five pump rates (40, 45, 50, 55, and 60
ml/min.). Dry, empty aluminum starch tables were initially tared
by suspending the table between two balances, one located on
each end of the starch table. The initial tared weight was the sum
of the weight measured by the two balances. Slurry was pumped
on the table at the designated flow rate and table slope, and the
overflow was recovered as the protein fraction. After all of the
slurry was pumped on the table, 150 ml of distilled water used to
rinse the bucket was pumped over the starch on the table to wash
the starch surface. The starch table was allowed to sit undisturbed
for at least 6 hr to air dry the starch, after which the starch table
was weighed across two balances to give a wet weight of starch.
Starch on the table was then immediately scraped off into tared
aluminum sample cups, and moisture was measured using a 2-hr
1350C forced-air oven procedure (AACC 1983). The overflow
(protein fraction) volume was measured, three 75-ml representative samples were placed in tared aluminum sample cups, and the
total solids content of the protein fraction was measured using a
two-stage forced-air oven procedure (AACC 1983). Protein in
starch (N x 6.25) was determined using a micro Kjeldahl proce-
dure (AOAC 1984).
As a control, duplicate 6,000-ml representative samples were
pumped on a 6.1-m x 8.3-cm table using the guide lines of the 1kg laboratory corn wet-milling procedure of Eckhoff et al (1993).
Slurry (6,000 ml) was chosen as representative of the volume of
slurry tabled during use of the 1-kg procedure. The table was set
at a slope of 0.0093 cm/cm with a pumping rate of 300 ml/min.
The percentage yield of starch, gluten, and the starch protein
content were determined.
RESULTS AND DISCUSSION
The average starch yield and protein content in the starch for
the 1-kg procedure controls was 80.0 and 0.52%, respectively.
This is =85% recovery of the starch in the starch-protein slurry.
This is a low recovery by commercial standards, but it may be
due to excessive recycling of middling material or other in-plant
situations. The protein content in the starch was at the high end of
values reported when using the 1-kg procedure. Eckhoff et al
(1993) reported a value of 0.32%; Tso and Eckhoff (1991) also
reported 0.32%. Singh and Eckhoff (1995) reported 0.42% protein in the starch they recovered using the 1-kg procedure, while
Anderson (1963), using a similar procedure, reported a value of
0.54%. The higher than desired protein content in the starch is
consistent with the lower starch recovery, indicating that the particular commercial slurry sample is difficult to separate.
Starch yield on the 100-g procedure table decreased with
increasing pump rate and table slope (Fig. 1). Holding slope or
pumping rate constant gave similar results in that at all pumping
rates tested the starch yield decreased with increasing table slope,
and at all table slopes tested, the starch yield decreased with
increasing pumping rate. The maximum starch yield of 88% was
at the lowest pumping rate and table slope (40 ml/min and 0.0052
cm/cm). The rate of starch yield decrease was greatest for the
lowest and highest table slopes (0.0052 and 0.0208 cm/cm) with
decreases of 6.9 and 7.3% starch yield. The middle two table
slopes (0.0104 and 0.0156 cm/cm) had decreases in starch yield
of 4.3 and 4.2%, respectively.
The protein content in the tabled starch also decreased with
increasing pump rate and table slope as anticipated (Fig. 2). Most
noticeable was that there was a large difference between the protein content of the starch tabled at 0.0052 cm/cm and the other
three table slopes, with significantly more protein being retained
on the table. There was no significant difference (a = 0.05)
between the protein content in the starch samples recovered at
rates greater than 50 ml/min except at the lowest table slope. Zipf
et al (1950) found a similar decreasing trend for the protein content in starch.
Selection of an appropriate table slope and pumping rate for
use with the 100-g laboratory wet-milling procedure was based
on conditions that give a protein content in starch and a starch
yield similar to what would be expected from tabling using the 1kg procedure. Protein content in starch levels >0.5% was not
considered acceptable, based upon industrial specifications. The
protein content in the starch decreased <0.5% at a table slopes
and pumping rates combinations >0.0104 cm/cm and 50 ml/min.
The best starch yield from the 100-g procedure for starch with
<0.5% protein was 81.8% at a 50 m/min pump speed and 0.0104
table slope, which was comparable to the values for the control
samples (80.0% starch yield and 0.52% protein in starch).
Although selection criteria for determining the table slope and
the pumping rate were based on comparison with results from the
1-kg procedure, there is no guarantee that numerical results will
always be highly correlated between the two procedures. Because
-
5 79
L-
u
a)
.076
1-40
1
I
45
50
Pumping Rate (ml/min.)
55
Fig. 1. Yield of starch as affected by table slope and slurry pumping rate.
52
CEREAL CHEMISTRY
40
45
60
w:,
I
50
Pumping Rate (ml/min.)
slurry pumping rate.
o
-A
starcn as attecteu
o
cf
rig. z. rioein colIIll oi
tawuea
55
1,.,
by
60
+-Ul-
-- A
taule siope anu
the table slopes are different (0.0104 cm/cm vs. 0.0093 cm/cm)
and the flow rate per cross-sectional area are not the same (10
ml/min/cm vs. 36 ml/mmn/cm), relative ranking of hybrids will
correlate between the two procedures, but numerical yield values
may be different.
CONCLUSIONS
Both starch yields and protein content in the starch decreased
with increasing table slope and pumping rate. The combination of
table slope and pumping rate that resulted in a starch yield and
starch protein content similar to that observed for the 1-kg tabling
procedure on the same slurry was 0.0104 cm/cm and 50 ml/min.
LITERATURE CITED
AMERICAN ASSOCIATION OF CEREAL CHEMISTS. 1983.
Approved Methods of the AACC, 8th ed. Method 44-18, Method 4419. The Association: St. Paul, MN.
AOAC. 1984. Official Methods of Analysis. Association of Official
Analytical Chemists. Method 7.025. AOAC: Washington, DC.
ANDERSON, R. A. 1957. A pilot plant for wet-milling. Cereal Science
Today 2(4):78-80.
ANDERSON, R. A. 1963. Wet milling properties of grains: Bench scale
study. Cereal Science Today 8(6):190-195, 221.
BLANCHARD, P. H. 1992. Technology of Corn Wet Milling and Associated Processes. Elsevier: Amsterdam.
BERKHOUT, F. 1976. The manufacture of maize starch. In: Starch Production Technology. J. A. Radley, ed. Applied Science Publishers:
London.
ECKHOFF, S. R., RAUSCH, K. D., FOX, E. J., TSO, C. C., WU, X.,
PAN, Z., and BURIAK, P. 1993. A laboratory wet-milling procedure
to increase reproducibility and accuracy of product yield. Cereal
Chem. 70:723-727.
ECKHOFF, S. R., SINGH, S. K., ZEHR, B. E., RAUSCH, K. D., FOX,
E. J., MISTRY, A. K., HAKEN, A. E., NIU, Y. X., ZOU, S. H.,
BURIAK, P., TUMBLESON, M. E., and KEELING, P. L. 1996. A
100-g laboratory corn wet-milling procedure. Cereal Chem. 73:54-57.
ECKHOFF, S. R., and TSO, C. C. 1991. Starch recovery from steeped
corn grits as affected by drying temperature and added commercial
protease. Cereal Chem. 68:319-320.
FOX, E. J., and ECKHOFF, S. R. 1993. Wet-milling of soft-endosperm,
high-lysine corn using short steep times. Cereal Chem. 70:402-404.
FOX, S. R., JOHNSON, L. A., HURBURGH, C. R., JR., DORSEYREDDING, C., and BAILEY, T. B. 1992. Relations of grain proximate
composition and physical properties to wet-milling characteristics of
maize. Cereal Chem. 69:191 -197.
HASSANEAN, A., and ABDEL-WAHED, A. 1986. A new method to
short the steeping period of corn grains. Starch/Staerke 38:417-419.
KERR, R. W. 1950. Chemistry and Industry of Starch. Academic Press:
Orlando, FL.
PELSHENKE, P. F., and LINDEMANN, E. 1954. Untershuchungen uber
die eignung verschiedener maissorten fur die starkefabrikation.
Staerke 6(8):177-182.
RAUSCH, K. D., SINGH, S. K., and ECKHOFF, S. R. 1993. Magnesium
bisulfite and sodium bisulfite as alternative steeping agents for wet
milling. Cereal Chem. 70:489-490.
RUBENS, R. W. 1990. A pilot plant for the wet milling of corn grain.
Cereal Foods World 35:1166-1169.
SHANDERA, D. L., PARKHURST, A. M., and JACKSON, D. S. 1995.
Interactions of sulfur dioxide, lactic acid, and temperature during
simulated corn wet milling. Cereal Chem. 72:371-378.
SINGH, N., and ECKHOFF, S. R. 1995. Hydrocyclone procedure for
starch-protein separation in laboratory wet milling. Cereal Chemistry
72:344-348.
STEINKE, J. D., and JOHNSON, L. A. 1991. Steeping maize in the
presence of multiple enzymes. I. Static batchwise steeping. Cereal
Chem. 68:7-12.
STEINKE, J. D., JOHNSON, L. A., and WANG, C. 1991. Steeping
maize in the presence of multiple enzymes. II. Continuous countercurrent steeping. Cereal Chem. 68:12-17.
WANG, C., and JOHNSON, L. A. 1992. Wet-milling characteristics of
propionate-treated high-moisture maize. I. Yields and composition of
products. Cereal Chem. 69:43-46.
WATSON, S. A., WILLIAMS, C. B., and WAKELY, R. D. 1951. Laboratory steeping used in a wet milling research program. Cereal Chem.
28:105.
WATSON, S. A. 1984. Corn and sorghum starches: Production. In:
Starch Chemistry and Technology. R. L. Whistler, J. N. BeMiller, and
E. F. Paschall, ed. Academic Press: Orlando, FL.
WEHLING, R. L., JACKSON, D. S., HOOPER, D. G., and GHAEDIAN,
A. R. 1993. Prediction of wet-milling starch yield from corn by nearinfrared spectroscopy. Cereal Chem. 70:720-723.
WELLER, C. L., PAULSEN, M. R., and STEINBERG, M. P. 1988. Correlation of starch recovery with assorted quality factors of four corn
hybrids. Cereal Chem. 65:392-397.
ZIPF, R. L., ANDERSON, R. A., and SLOTTER, R. L. 1950. Wet milling of grain sorghum. Cereal Chem. 27:463-473.
[Received September 14, 1995. Accepted September 28, 1995.]
Vol. 73, No. 1, 1996
53